EF1 tornado, Glazier to Arnett, Texas, May 18, 2025

1. EF065 to 85 mph

   Light damage. Surface peeled off some roofs, gutters and siding damaged, branches broken, shallow-rooted trees pushed over.

2. EF186 to 110 mphthis tornado

   Moderate damage. Roofs badly stripped, mobile homes overturned or badly damaged, exterior doors lost, windows broken.

3. EF2111 to 135 mph

   Considerable damage. Roofs torn off well-built houses, frame homes shifted on their foundations, mobile homes destroyed, large trees snapped or uprooted, cars lifted off the ground.

4. EF3136 to 165 mph

   Severe damage. Whole stories of well-built houses destroyed, large buildings badly damaged, trains overturned, trees debarked, heavy cars lifted and thrown.

5. EF4166 to 200 mph

   Devastating damage. Well-built houses leveled, structures with weak foundations moved some distance, cars thrown and large missiles generated.

6. EF5over 200 mph

   Incredible damage. Strong frame houses swept off their foundations, car-sized missiles thrown more than 100 meters, steel-reinforced concrete badly damaged, high-rise buildings deformed.

7. Unrated

   The survey recorded a tornado but could not put a number on it, most often because it crossed open country and damaged nothing a rating can be read from.

Rating

EF1

Date

May 18, 2025

Touchdown

5:21 PM CDT (22:21 UTC)

Lifted

5:27 PM CDT (22:27 UTC)

On the ground

6 minutes

Path length

1.7 miles (2.7 km)

Path width

75 yards (69 m)

Deaths

0

Injuries

0

Property damage

$35,000

Counties

Hemphill and Ellis counties

Touchdown point

36.009, -100.005

Lift-off point

36.028, -99.986

Record number

2505181621-01

51 damage points were recorded along this path by NWS OUN, in the Damage Assessment Toolkit the survey teams carry. The strongest indicator was rated EF3. The highest wind the damage implies is 143 mph.

• Very large tree with expansive and healthy root system
• Numerous live hardwoods shredded with few limbs remaining. Piles of limbs collected in barbed wire fence across the road.
• Half of tree tossed across the street? Unclear if the stump and 1/2 tree are the same or not. Half tree did not originate where it is now.
• Ground scouring evident from road. Crops scoured to bare soil.

Survey points are preliminary and are the raw material the published rating was built from. Source: NWS Damage Assessment Toolkit.

In the late afternoon to early evening hours of May 18, a strong upper level storm system and surface dryline created favorable conditions for severe thunderstorms across the far eastern Panhandles. This environment was characterized by extreme instability and very high surface dewpoints in the low 70s east of the dryline, which supported all hazards of severe weather. Two supercells developed in the eastern Texas Panhandle. One moved to the north and produced large hail up to golfball size before moving into the Oklahoma Panhandle. The other also produced large hail up to baseball size along with a couple of tornadoes near the state line with western Oklahoma.

A deep upper-level trough continued to progress eastward across the Central and Southern Plains on the 18th, promoting another round of intense severe weather across the northwestern portion of the WFO Norman Forecast Area. During the early evening, a strong dryline served as the focus for thunderstorm initiation across the aforementioned area. A pair of intense supercell thunderstorms developed here, each producing large to very large hail, with multiple reports of hailstones ranging from four to five inches in diameter. In addition to very large hail, one of the supercells generated a tornado family consisting of four separate tornadoes across Ellis County. Among these was an intense EF3 tornado that impacted areas just southwest of Arnett, OK, causing significant damage along its path.

The tornado started near County Road 30 on the very eastern edge of Hemphill County and quickly intensified as it crossed onto property where it produced damage to several buildings and a home. The home sustained some roofing damage and the back corner of the house had a partial uplift of the roof deck. Additional storage buildings were damaged including shed doors blown off and roofs and siding damaged. Several trees were uprooted in the yard. A cattle wind break was partially torn from the ground. The tornado then crossed into Ellis County Oklahoma (NWS Norman forecast area), where it snapped a few tree limbs on E690. The tornado then crossed an open field and the last noted damage was along NS1690 where some large branches were snapped. Based on storm spotter footage and survey, it appears the tornado lifted shortly thereafter. The tornado was rated an EF1 based on the house damage and the tree damage. The entire track length was around 1.6 miles. The maximum estimated winds were 95mph.

This tornado moved into Ellis County from Hemphill County TX crossing the state line just south of Texas Farm-to-Market Road 2758 and Oklahoma County Road E0690. A storm survey by the NWS Office in Amarillo observed tree damage along E0690 Road just east of the Texas state line, and along County Road N1690 south of E0680 Road. The tornado dissipated very closer to E0680 Road about 250 yards west of N1690 Road.||Geolocation of drone footage of the tornado from the OTUS (Observations of Tornadoes by UAV Systems) Research team showed that while the tornado moved generally northeast, numerous variations in the motion occurred.||This tornado was rated EF1 based on damage in Hemphill County TX, but only EF0 damage could be confirmed in Oklahoma.

Immediately after the EF3 tornado dissipated west of Arnett, video from an off-duty Storm Prediction Center employee observed an anticyclonic tornado looking south from the intersection of County Roads E0620 and N1810. The tornado was brief, and the specific location is estimated.

After the EF3 tornado dissipated just west of Arnett, another tornado developed just northwest of Arnett and moved northeast. Trees and an oil field building were damaged by this tornado.

From NOAA NCEI's Storm Events Database, the narrative the local forecast office wrote for Storm Data.

9 minutes of warning. NWS AMA issued a tornado warning for this county at 22:12 UTC, 9 minutes before the tornado touched down, in force until 22:39 UTC.

The county was under a tornado watch, issued at 20:41 UTC and running to 01:11 UTC.

Every warning and watch covering the touchdown point that day, from the Iowa Environmental Mesonet's archive of National Weather Service products. Lead time is measured from the issue of the warning in force to the touchdown time in the database, both of which were recorded to the minute by different people.

The Storm Prediction Center had this point in a Slight Risk, with a 0.05 chance of a tornado within 25 miles. The outlook in force was issued at 20:05 UTC.

SPC's Day 1 convective outlook as it stood when the tornado began, read at the touchdown point. The same archive is on the radar archive.

The radiosonde launches on either side of the tornado, from the nearest station that was launching then, drawn and analyzed the way Storm Lab does: the same diagram, the same parcel and shear definitions SPC's mesoanalysis uses.

Gage/Shattuck (GAG), 37 km from the path, in the two hours either side.

From the Iowa Environmental Mesonet's archive of airport observations.

The Amarillo radar (KAMA), 178 km away, was scanning. The volume nearest the touchdown began at 22:19 UTC.

Open the radar archive at this moment, over the track

Reflectivity and velocity, decoded from the Level II archive in the browser. Volumes take a few seconds to open the first time.

Local storm reports within 60 km of the path, in the three hours either side.

From the Iowa Environmental Mesonet's archive of National Weather Service local storm reports.

The live radar over the trackEvery 2025 tornado in TexasEvery tornado in Texas since 1950Warnings in force in Texas nowThe forecast where you are

42 more were confirmed in the United States that day.

Ratings are the survey's, read off the damage rather than measured; the rating scale is above. Tornado counts rise through the record as reporting improved, so a year in the 1950s is not comparable with one today except for the strongest tornadoes, which were always noticed.

NOAA Storm Prediction Center tornado database, 1950-2025, the official record kept by NOAA's Storm Prediction Center.